A MUTATIONAL MODEL FOR CHILDHOOD CANCER Louise C. Strong, M.D. The overall goal of the program is to identify genes that predispose to childhood cancer, the molecular pathways to tumor development, and the clinical implications. We have focused on two model familial syndromes of childhood and adolescent cancers, sarcomas and Li Fraumeni syndrome (LFS) and its variants, and Wilms' tumor of the kidney. We have developed a multidisciplinary program to investigate genetic susceptibility to childhood and associated cancer using integrated technology of genetic epidemiology, molecular genetics and genomics applied to the rich resources of cancer prone families and mouse models developed in this program. The hypotheses are based on a multi-stage model for cancer. For each tumor type, genetic loci have been identified that may be altered both as germline mutations and as tumor-specific mutations. There is also significant evidence for additional cancer susceptibility genes and risk modifiers, including an effect of generation, at least for LFS. The underlying themes of the program include identification of the underlying cancer susceptibility genes and risk modifiers, analysis of germline and somatic mutations by type and mechanism, determination of the molecular genetic anatomy of the tumors, development of animal models for human cancer susceptibility syndromes, determination of the role of telomere function in cancer risk in LFS and mouse models, and determination of the implications of germline mutations for the patients and their families. The findings from this program should provide insights into the mechanisms of carcinogenesis as well as guidelines for clinical programs for patients at high risk of cancer.